Patient data: finding the right balance

You might have missed it. On the day the Royal College of Physicians released its much welcomed report of the Future Hospital Commission, and of the first NHS England AGM, the government published its response to the second Caldicott review. Short of the arcane rules of payment by results, information governance is perhaps one of the most complex and difficult non-clinical subjects in the running of the NHS.

The Caldicott review, published in April, aimed to help make things clearer and strike the right balance between the protection of patient information and the use of that information to improve patient care. In its response, the government has accepted, in principle, all of the recommendations of the review, seeking to follow not just the 26 recommendations but the spirit of the review as a whole.

Andrew Fenton

‘There is a view that in reality the balance has been tipped further in the direction of locking down information that is essential to the improvement of patient care’

So, has it got it right? Has a good balance been struck? Has the information governance landscape become clearer? It is probably an impossible task, but the endeavours to openly tackle this difficult area should be welcomed. There is much in the review and its acceptance by government that sets out important principles and directions for practical application.

But on the ground in commissioning land, there is much disquiet and frustration about the new information governance arrangements resulting from the Health and Social Care Act and by implication the recommendations of the Caldicott review. There is a view that in reality the balance has been tipped further in the direction of locking down information that is essential to the improvement of patient care.

A letter from NHS Clinical Commissioners in June set out a list of concerns about the constraints over the use of patient information. These concerns relate in part to the transactional nature of the purchaser-provider split (such as invoice validation), but there are even more important issues about how we use data to improve care and the integration of local health economies.

Care continuum

At the heart of the concerns is the sense that clinically led health system reform has not been well understood by the review process, and even that the underlying legislative framework is not conducive to the challenges set on commissioners and their local health systems.

While the review recommendations make it clear that information should be shared appropriately for direct patient care between providers, it makes a simple binary distinction between direct patient care and secondary uses of data, such as planning and commissioning. In reality, in a clinically led commissioning environment, it is more of a continuum between direct patient care and commissioning. An example can help illustrate this.

‘There are indications in the Caldicott review that the exchange of information across providers to support direct patient care and better integration of services could become harder’

GPs, as direct providers of patient care, need to know about patients that have or are likely to be hitting urgent care services: they need identified patient-level data to plan and provide upstream care for those patients. The same information is needed (in anonymised or pseudonymised form) to enable a population level view of patient need, stratified by level and type of patient need.

While this requirement is fully recognised by the review, the recent practical application guidance results in a convoluted and costly set of technical models designed to ensure legal compliance while attempting to deliver a useable end product for commissioners and GPs as providers. There is a risk that implementing such solutions will be put in the “too hard” tray, and GPs and commissioners in some areas will not have the benefit of flexible systems to help with urgent care demand. 

Artificial concept

Beyond the issues around risk stratification and real time information on urgent care demand, there are also indications in the Caldicott review that the exchange of information across providers to support direct patient care and better integration of services could become harder.

This sounds perverse, as the intention and claim of the review is that sharing of information for patient care should become easier, no longer bound by the old “IG says no” approach to information sharing. However, where the review talks about information sharing for patient care, it is largely focused on individual bilateral messages and communications between providers − for example, relating to a referral, a discharge, a lab result or a care plan.

What this misses is that the ambition of more joined up, patient focused health systems requires not fragmented and complex sets of patient data flying between providers (including GPs), but a more advanced and holistic approach to electronic patient record integration. The development of “health information exchanges” or virtual shared care records within local health communities is a key enabler to the ambitions of integrated care.

The review, however, indicates that while implied consent can be taken for the individual communications between providers, sharing of a wider set of information could require explicit consent − an unattainable goal for any viable patient record system.

‘At a time when we are working towards more holistic, system wide and patient focused delivery of care, the management and use of information is at risk of being fragmented’

The government response on this recommendation puts the onus on professional regulators to produce consistent guidance for when implied consent is acceptable. In addition, there needs to be work to define the scope of a core shared patient record that can support local interoperability developments. Some health systems have already done this via local consultation in the rollout of local summary care records. The national summary care record on its own does not have the degree of clinical content to enable effective integrated care management across providers, and only contains information from GP clinical systems.

The challenges relating to risk stratification and local care records are largely because they involve “record linkage”, in that patient information from multiple systems is combined to create a patient and population level view. Record linkage is seen as requiring an extra level of control − a sound principle. But it is a rather artificial concept, an artefact of separate providers with different information systems.

At a time when we are working towards more holistic, system wide and patient focused delivery of care, the management and use of information is at risk of being fragmented or exchanged only through isolated communications between providers. The new information governance landscape taking shape risks sitting uneasily with the aim of properly integrating and improving care across our health and social care communities.

Andrew Fenton is an associate director of commissioning support at Central Southern CSU